Pith. sign in

REVIEW 5 major objections 5 minor 94 references

Source-free Semantic Regularization Learning for Semi-supervised Domain Adaptation

T0 review · 5 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A source-free fine-tuning scheme that freezes the classifier and applies three semantic regularization losses claims large accuracy gains over prior semi-supervised domain adaptation methods on three benchmarks.

desk verdict A coherent source-free SSDA recipe whose headline gains are not yet reproducible from the paper as written; worth refereeing, not worth citing yet. read the letter →

arxiv 2501.01126 v1 pith:E2GQH62W submitted 2025-01-02 cs.CV

Xinyang Huang , Chuang Zhu , Ruiying Ren , Shengjie Liu , Tiejun Huang This is my paper · ORCID
classification cs.CV
keywords semi-superviseddomainadaptationsource-freesemanticregularizationcontrastivelearningmixuppseudo-labelnoisetargetprediction
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that semi-supervised domain adaptation can be substantially improved by treating the target domain as a source-free fine-tuning problem and regularizing the feature extractor with semantic information extracted from target probabilities. It proposes SERL, which combines three losses: semantic probability contrastive regularization, hard-sample mixup regularization, and target prediction regularization. On DomainNet, Office-Home, and Office-31, SERL reports average accuracy gains of roughly 11 percentage points over the previous best source-free method DEEM, and larger gains over source-access baselines. The claim matters because it suggests that a frozen source classifier plus carefully chosen regularization can capture richer target semantics than alignment-based adaptation.

What carries the argument

The load-bearing mechanism is the frozen classifier's weight vectors used as category anchors. SPCR operates on probability outputs, pulling predictions of the same pseudo-class together and pushing others apart, with adaptive weights that down-weight low-confidence pairs. HMR picks the nearest and farthest unlabeled samples from each anchor to form easy and hard sets, then interpolates them with mixup under an L2 loss. TPR applies early-learning regularization, in which a moving average of past predictions constrains current outputs, reducing the influence of erroneous pseudo-labels. KNN-based pseudo-label propagation inherited from DEEM supplies the base labels that the regularizers build on.

What would settle it

Measure the agreement between the frozen-anchor easy/hard partition and the true categories on a target domain with substantial shift: if the nearest-to-anchor samples are not predominantly the same class as the anchor, the HMR mixing step is mixing unrelated classes and the mechanism described in the paper is not what produces the reported accuracy.

Watch

Extended reading notes

Core claim

SERL's central discovery is that freezing the source classifier and fine-tuning only the feature extractor on the target domain, while imposing three semantic regularizers, yields state-of-the-art SSDA performance without any access to source data during adaptation. The three regularizers are SPCR, which applies a contrastive loss to predicted probabilities with confidence-based adaptive weights; HMR, which separates target samples into easy and hard sets by cosine distance to frozen classifier weight vectors and mixes them; and TPR, which anchors current predictions to a moving average of early predictions to resist pseudo-label noise. The reported numbers: DomainNet average 89.9% (1-shot) and 90.9% (3-shot), improving DEEM by 11.8 and 11.4 points; Office-Home average 80.6% and 83.5%, improving DEEM by 10.1 and 10.6 points; Office-31 average 80.1% and 82.3%, improving DEEM by 3.9 and 4.6 points.

Load-bearing premise

The method assumes that the frozen source classifier's weight vectors still point at the correct category centers on the target domain, so that the easy/hard split and the pseudo-labels built around them are meaningfully aligned with target semantics.

Editorial extensions

If this is right

  • If correct, source-free SSDA no longer needs to trade away accuracy; a frozen classifier can serve as a stable semantic anchor during target fine-tuning.
  • The large margins on DomainNet and Office-Home suggest that regularization on target semantics alone, without source-access or adversarial alignment, can outperform source-with methods on standard benchmarks.
  • The easy/hard sample split, if it works as described, makes low-confidence unlabeled target samples a source of learning signal rather than a liability.
  • The early-prediction anchoring (TPR) implies that pseudo-label noise can be countered by remembering the model's own early predictions rather than by re-weighting or filtering noisy labels.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural next test is whether the same three regularizers transfer to open-set or partial-set SSDA, where a frozen classifier's anchors may not line up with target categories.
  • One could isolate HMR's contribution by comparing the reported anchor-based easy/hard split against a random split using the same mixup loss; a large gap would confirm that difficulty selection, not mixup alone, drives the gains.
  • Because SPCR operates on softmax outputs rather than features, it may behave as a confidence-shaping regularizer that could also benefit fully unsupervised domain adaptation, though the paper does not test that setting.
  • An independent re-implementation with identical backbones, batch sizes, and training schedules would clarify how much of the 11-point gap is due to the regularizers rather than to source-free initialization differences.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper proposes SERL, a source-free semi-supervised domain adaptation (SSDA) framework that fine-tunes a source-pretrained feature extractor while freezing the classifier. Three regularization terms are introduced: semantic probability contrastive regularization (SPCR), which contrasts probability outputs with adaptive weights; hard-sample mixup regularization (HMR), which mixes easy and hard target samples selected by distance to frozen classifier anchors; and target prediction regularization (TPR), which uses early-learning moving-average predictions to reduce the influence of noisy pseudo-labels. The method is evaluated on DomainNet, Office-Home, and Office-31 in 1-shot and 3-shot settings, with reported average gains over the prior source-free method DEEM of 11.8/11.4 points on DomainNet, 10.1/10.6 points on Office-Home, and 3.9/4.6 points on Office-31.

Significance. If the reported results are reproducible, SERL would be a substantial advance in source-free SSDA, a practically important setting where target labels are scarce and source data is unavailable during adaptation. The paper includes component ablations (Table VI), sensitivity analyses (Figures 5–6), three random seeds, comparisons with many baselines, and qualitative analyses. The core idea is coherent and each component contributes positively in the reported ablations. However, the central state-of-the-art claim is not currently independently checkable: key protocol details are missing or internally inconsistent, hyperparameters appear to be tuned on the same evaluation tasks, and no code or per-seed variance is provided. The method's reliance on pseudo-labels and frozen classifier anchors also creates a real risk under large domain shifts, as acknowledged by the paper's own emphasis on hard samples.

major comments (5)
  1. [Section IV-B and IV-E.1] The loss weight λprob is given as 0.3 in the implementation details (Section IV-B) but as 0.1 in the sensitivity analysis (Section IV-E.1, Figure 5), and the paper never states which value was used for Tables I–VII. Because the sensitivity curve shows nontrivial accuracy changes with λprob, the reported SOTA numbers cannot be attributed to the described method unless this contradiction is resolved and the exact setting is specified.
  2. [Algorithm 1 and Section IV-B] The number of training epochs T in Algorithm 1 is never specified, and the KNN-based pseudo-label propagation 'following [67]' omits the number of neighbors k, the low-uncertainty threshold, the feature space used for the graph, and the update frequency. These choices can materially affect accuracy, so the method is not reproducible as written.
  3. [Section III-C, Eqs. (11)-(13)] The easy/hard selection in Eqs. (11)–(12) uses only cosine distance to anchor c without requiring that the selected sample's pseudo-label be c, and Eq. (13) does not specify how X_easy_i pairs with X_hard_j or where y_easy and y_hard come from. A literal implementation can therefore mix samples from different classes, so the claimed within-class guidance of HMR is not guaranteed by the stated equations.
  4. [Section III-A and Figure 3] Figure 3 shows weak and strong augmentations for unlabeled data, but Section IV-B defines only RandAugment as the strong augmentation and never defines the weak augmentation. It is also unclear which view (original, weak, or strong) enters Eqs. (4), (8), and (14), making the exact training objective ambiguous.
  5. [Tables I-V and Figures 5-6] The reported gains are large, yet no per-seed standard deviations are given despite the statement that three random seeds were used, and the sensitivity analysis in Figures 5–6 tunes hyperparameters on evaluation scenarios that also appear in Tables I–V without a separate validation split. Please report mean±std over seeds and clarify whether the main tables use the Section IV-B defaults or the optimal sensitivity-analysis values, since this is essential for assessing possible selection bias.
minor comments (5)
  1. [Section IV-A] The description of DomainNet says the dataset is 'composed of 345 classes' and then states 'each domain contains 126 image categories'; this is confusing and should be clarified to reflect the standard 126-class subset used by MME.
  2. [Section IV-B] There is a typo in 'Randaugmnt' which should be 'RandAugment'.
  3. [Section III-D, Eq. (15)] The abstract and introduction describe TPR as 'maximizing the correlation between the current prediction and the past learned objective,' but Eq. (15) is a log-based early-learning regularization term; the connection between these descriptions should be stated more precisely.
  4. [Section III-B, Eq. (10)] The equivalence in Eq. (10) is correct only under the assumption that probability vectors are normalized and nonnegative; this is true by definition, but the statement would be clearer if the assumption were acknowledged.
  5. [Section IV-E.1 and Figure 5] The text says the sensitivity analysis is conducted on the Office-Home C→A scenario, while the caption of Figure 5 also mentions DomainNet R→C; please clarify whether both scenarios are shown or whether the description is incomplete.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity: the central claim is an empirical benchmark result, and the regularizers are self-referential in the standard semi-supervised sense but are not fitted to or derived from the reported accuracy. Minor same-group citations are present but do not support the SOTA claim.

full rationale

The paper's central claim is empirical: SERL improves on DEEM by large margins on DomainNet, Office-Home, and Office-31. The derivation chain is a training objective composed of three regularizers applied to a frozen-classifier, source-free fine-tuning setup. None of the losses is constructed from the benchmark accuracies or from DEEM's outputs; the reported numbers are measured on held-out target labels after training. The self-referential element is that pseudo-labels and adaptive contrastive weights are generated by the model being adapted (e.g., Eq. 4 uses argmax of the model's own probability output, Eq. 9 defines the SPCR weight as the dot product of the model's own probabilities, and Eq. 13 uses pseudo-label-derived mixup targets). This is standard semi-supervised self-training behavior, not a circular derivation of the empirical result: the evaluation is external to those training signals. The same-group citations are minor and not load-bearing: [33] is cited for the choice of strong augmentation, and [75] is cited for a motivation statement about noisy pseudo-labels and clustering structure. Neither citation is used to justify the claimed improvement, to import a uniqueness theorem, or to supply a fitted value. A separate reproducibility concern exists, but it is not circularity: Section IV-B fixes lambda_prob = 0.3 while Section IV-E.1 and Figure 5 report the best lambda_prob as 0.1; the number of epochs T, the weak-augmentation definition, and KNN propagation details are under-specified. These omissions affect whether the +11.8/+11.4-point claims can be independently verified, but they do not make the derivation equivalent to its inputs by construction.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

No new physical or formal entities are introduced; the paper's contributions are loss functions and a training recipe. The free parameters are loss weights, temperature, mixup shape, sample counts, and inherited KNN details, all chosen by hand, by sensitivity scans, or by referencing prior work.

free parameters (8)
  • lambda_prob = 0.3
    Loss weight for SPCR; selected by sensitivity analysis on Office-Home C->A and DomainNet R->C (Fig. 5).
  • lambda_mix = 60
    Loss weight for HMR; selected in the same sensitivity analysis (Fig. 5).
  • lambda_pre = 3
    Loss weight for TPR; selected in the sensitivity analysis (Fig. 5).
  • beta = 0.7
    Moving average coefficient in target prediction regularization (Section III-D); chosen without a dedicated sensitivity analysis.
  • tau = 0.15
    Temperature in the contrastive loss (Eq. 7); inherited from prior contrastive learning.
  • alpha = 1.0
    Beta distribution parameter for mixup (Eq. 13); standard choice from the mixup literature.
  • N_easy and N_hard = 15
    Number of easy/hard target samples selected per class; tuned in Fig. 6.
  • KNN label propagation parameters = not specified
    Neighbor count, confidence threshold, and propagation schedule are inherited from DEEM [67] but not reported, making the method under-specified.
assumptions (5)
  • standard math InfoNCE contrastive loss and probability dot products provide a valid learning signal for target feature discrimination.
    Section III-B uses Eqs. 7-10 as the backbone of SPCR.
  • domain assumption The frozen source classifier remains a reliable decision boundary in the target domain.
    Section III-A: 'Following [38], [67], we freeze f(.) and train g(.)' during adaptation.
  • domain assumption Pseudo-labels generated by the model and KNN propagation are sufficiently accurate to serve as supervision.
    Eq. 4 uses argmax predictions as labels for L_u, and Eq. 9 uses them to define positive contrastive pairs.
  • domain assumption Source classifier weight vectors can serve as semantic anchors for easy/hard sample selection.
    Section III-C, Eqs. 11-12; based on neural collapse results [81].
  • domain assumption Early-training predictions are more reliable than later predictions in the presence of noisy pseudo-labels.
    Section III-D, Eq. 15, borrowed from ELR [44].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Source-free Semantic Regularization Learning for Semi-supervised Domain Adaptation." pith.science (2026). https://pith.science/paper/E2GQH62W

@misc{pith2026250101126,
  author       = {Pith},
  title        = {Pith review of: Source-free Semantic Regularization Learning for Semi-supervised Domain Adaptation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E2GQH62W}},
  note         = {Machine review of arXiv:2501.01126}
}
read the original abstract

Semi-supervised domain adaptation (SSDA) has been extensively researched due to its ability to improve classification performance and generalization ability of models by using a small amount of labeled data on the target domain. However, existing methods cannot effectively adapt to the target domain due to difficulty in fully learning rich and complex target semantic information and relationships. In this paper, we propose a novel SSDA learning framework called semantic regularization learning (SERL), which captures the target semantic information from multiple perspectives of regularization learning to achieve adaptive fine-tuning of the source pre-trained model on the target domain. SERL includes three robust semantic regularization techniques. Firstly, semantic probability contrastive regularization (SPCR) helps the model learn more discriminative feature representations from a probabilistic perspective, using semantic information on the target domain to understand the similarities and differences between samples. Additionally, adaptive weights in SPCR can help the model learn the semantic distribution correctly through the probabilities of different samples. To further comprehensively understand the target semantic distribution, we introduce hard-sample mixup regularization (HMR), which uses easy samples as guidance to mine the latent target knowledge contained in hard samples, thereby learning more complete and complex target semantic knowledge. Finally, target prediction regularization (TPR) regularizes the target predictions of the model by maximizing the correlation between the current prediction and the past learned objective, thereby mitigating the misleading of semantic information caused by erroneous pseudo-labels. Extensive experiments on three benchmark datasets demonstrate that our SERL method achieves state-of-the-art performance.

Figures

Figures reproduced from arXiv: 2501.01126 by the authors.

Figure 1
Figure 1. The learning scenario of our SERL framework. Different from [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The motivation of our SERL. (a) Due to the scarcity of target semantic [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Illustration of our proposed semantic regularization learning (SERL) framework. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: The impact of source-free learning frameworks on performance. The [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: The effect of different loss balance parameters [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 8
Figure 8. Figure 8: The SVD analysis of feature matrices obtained by different methods [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 7
Figure 7. Figure 7: Histogram of quantitative comparison under different number of [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 9
Figure 9. Figure 9: Feature visualization using t-SNE [91]. We randomly selected seven categories and assigned them different colors for the 3-shot scenes of DomainNet R→S and Office-Home C→P. The red box shows obvious differences. (a) Easy Samples (b) Hard Samples [PITH_FULL_IMAGE:figur…
Figure 10
Figure 10. Figure 10: The Grad-CAM [92] visualization of the features generated by our SERL for different samples in the DomainNet dataset. smaller singular values positively affect the classification and intuitively improve the discriminability of the features. F. Feature Visualization 1)…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

94 extracted references · 61 canonical work pages

  1. [67]

    Context-guided entropy minimization for semi-supervised domain adaptation,

    N. Ma, J. Bu, L. Lu, J. Wen, S. Zhou, Z. Zhang, J. Gu, H. Li, and X. Yan, “Context-guided entropy minimization for semi-supervised domain adaptation,” Neural Networks, vol. 154, pp. 270–282, 2022

    work page 2022

  2. [1]

    Imagenet classification with deep convolutional neural networks,

    A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,” Advances in neural informa- tion processing systems , vol. 25, 2012

    2012

  3. [2]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , pp. 770–778, 2016

    2016

  4. [3]

    Xnor-net: Imagenet classification using binary convolutional neural networks,

    M. Rastegari, V . Ordonez, J. Redmon, and A. Farhadi, “Xnor-net: Imagenet classification using binary convolutional neural networks,” in European conference on computer vision , pp. 525–542, Springer, 2016

    2016

  5. [4]

    Imagenet classification with deep convolutional neural networks,

    A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,” Communications of the ACM, vol. 60, no. 6, pp. 84–90, 2017

    2017

  6. [5]

    Transductive episodic-wise adaptive metric for few-shot learning,

    L. Qiao, Y . Shi, J. Li, Y . Wang, T. Huang, and Y . Tian, “Transductive episodic-wise adaptive metric for few-shot learning,” in Proceedings of the IEEE/CVF international conference on computer vision , pp. 3603– 3612, 2019

    2019

  7. [6]

    Image classification by cross-media active learning with privileged information,

    Y . Yan, F. Nie, W. Li, C. Gao, Y . Yang, and D. Xu, “Image classification by cross-media active learning with privileged information,” IEEE Transactions on Multimedia , vol. 18, no. 12, pp. 2494–2502, 2016

    2016

  8. [7]

    Csps: An adaptive pooling method for image classification,

    J. Wang, W. Wang, R. Wang, and W. Gao, “Csps: An adaptive pooling method for image classification,” IEEE Transactions on Multimedia , vol. 18, no. 6, pp. 1000–1010, 2016

    2016

Show all 94 references
  1. [8]

    Fully convolutional networks for semantic segmentation,

    J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in Proceedings of the IEEE conference on computer vision and pattern recognition , pp. 3431–3440, 2015

    2015

  2. [9]

    Fbsnet: A fast bilateral symmetrical network for real-time semantic segmentation,

    G. Gao, G. Xu, J. Li, Y . Yu, H. Lu, and J. Yang, “Fbsnet: A fast bilateral symmetrical network for real-time semantic segmentation,” IEEE Transactions on Multimedia , 2022. 12

    2022

  3. [10]

    Semantic segmentation guided pixel fusion for image retargeting,

    B. Yan, X. Niu, B. Bare, and W. Tan, “Semantic segmentation guided pixel fusion for image retargeting,” IEEE Transactions on Multimedia , vol. 22, no. 3, pp. 676–687, 2019

    2019

  4. [11]

    Image segmentation using deep learning: A survey,

    S. Minaee, Y . Boykov, F. Porikli, A. Plaza, N. Kehtarnavaz, and D. Terzopoulos, “Image segmentation using deep learning: A survey,” IEEE transactions on pattern analysis and machine intelligence, vol. 44, no. 7, pp. 3523–3542, 2021

    2021

  5. [12]

    Muva: A new large-scale benchmark for multi-view amodal instance segmentation in the shopping scenario,

    Z. Li, W. Ye, J. Terven, Z. Bennett, Y . Zheng, T. Jiang, and T. Huang, “Muva: A new large-scale benchmark for multi-view amodal instance segmentation in the shopping scenario,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. 23504–23513, 2023

    2023

  6. [13]

    Seggpt: Segmenting everything in context,

    X. Wang, X. Zhang, Y . Cao, W. Wang, C. Shen, and T. Huang, “Seggpt: Segmenting everything in context,” arXiv preprint arXiv:2304.03284 , 2023

    2023 arXiv

  7. [14]

    Domain adaptation via transfer component analysis,

    S. J. Pan, I. W. Tsang, J. T. Kwok, and Q. Yang, “Domain adaptation via transfer component analysis,” IEEE transactions on neural networks, vol. 22, no. 2, pp. 199–210, 2010

    2010

  8. [15]

    Visual domain adaptation: A survey of recent advances,

    V . M. Patel, R. Gopalan, R. Li, and R. Chellappa, “Visual domain adaptation: A survey of recent advances,” IEEE signal processing magazine, vol. 32, no. 3, pp. 53–69, 2015

    2015

  9. [16]

    Instance adaptive self-training for unsupervised domain adaptation,

    K. Mei, C. Zhu, J. Zou, and S. Zhang, “Instance adaptive self-training for unsupervised domain adaptation,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXVI 16, pp. 415–430, Springer, 2020

    2020

  10. [17]

    Self-guided adaptation: Progressive representation alignment for do- main adaptive object detection,

    C. Zhang, Z. Li, J. Liu, P. Peng, Q. Ye, S. Lu, T. Huang, and Y . Tian, “Self-guided adaptation: Progressive representation alignment for do- main adaptive object detection,” IEEE Transactions on Multimedia , vol. 24, pp. 2246–2258, 2021

    2021

  11. [18]

    Unsupervised domain adaptation by backpropagation,

    Y . Ganin and V . Lempitsky, “Unsupervised domain adaptation by backpropagation,” in International conference on machine learning , pp. 1180–1189, PMLR, 2015

    2015

  12. [19]

    Informative feature disentanglement for unsupervised do- main adaptation,

    W. Deng, L. Zhao, Q. Liao, D. Guo, G. Kuang, D. Hu, M. Pietik ¨ainen, and L. Liu, “Informative feature disentanglement for unsupervised do- main adaptation,” IEEE Transactions on Multimedia, vol. 24, pp. 2407– 2421, 2021

    2021

  13. [20]

    Cross- domain contrastive learning for unsupervised domain adaptation,

    R. Wang, Z. Wu, Z. Weng, J. Chen, G.-J. Qi, and Y .-G. Jiang, “Cross- domain contrastive learning for unsupervised domain adaptation,” IEEE Transactions on Multimedia , 2022

    2022

  14. [21]

    Adversarial mixup ratio confusion for unsupervised domain adaptation,

    M. Jing, L. Meng, J. Li, L. Zhu, and H. T. Shen, “Adversarial mixup ratio confusion for unsupervised domain adaptation,” IEEE Transactions on Multimedia, 2022

    2022

  15. [22]

    Discriminative invariant alignment for unsupervised domain adaptation,

    Y . Lu, D. Li, W. Wang, Z. Lai, J. Zhou, and X. Li, “Discriminative invariant alignment for unsupervised domain adaptation,” IEEE Trans- actions on Multimedia , vol. 24, pp. 1871–1882, 2021

    2021

  16. [23]

    A review of single-source deep unsupervised visual domain adaptation,

    S. Zhao, X. Yue, S. Zhang, B. Li, H. Zhao, B. Wu, R. Krishna, J. E. Gonzalez, A. L. Sangiovanni-Vincentelli, S. A. Seshia, et al., “A review of single-source deep unsupervised visual domain adaptation,” IEEE Transactions on Neural Networks and Learning Systems , vol. 33, no. 2...

    2020

  17. [24]

    Dual structural knowledge interaction for domain adaptation,

    Y . Zuo, H. Yao, L. Zhuang, and C. Xu, “Dual structural knowledge interaction for domain adaptation,” IEEE Transactions on Multimedia , no. 99, pp. 1–15, 2023

    2023

  18. [25]

    Unsupervised domain adaptation via risk-consistent estimators,

    F. Ding, J. Li, W. Tian, S. Zhang, and W. Yuan, “Unsupervised domain adaptation via risk-consistent estimators,” IEEE Transactions on Multimedia, 2023

    2023

  19. [26]

    Semi- supervised domain adaptation via minimax entropy,

    K. Saito, D. Kim, S. Sclaroff, T. Darrell, and K. Saenko, “Semi- supervised domain adaptation via minimax entropy,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. 8050– 8058, 2019

    2019

  20. [27]

    Cross-domain adaptive clustering for semi-supervised domain adaptation,

    J. Li, G. Li, Y . Shi, and Y . Yu, “Cross-domain adaptive clustering for semi-supervised domain adaptation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 2505– 2514, 2021

    2021

  21. [28]

    Ecacl: A holistic framework for semi-supervised domain adaptation,

    K. Li, C. Liu, H. Zhao, Y . Zhang, and Y . Fu, “Ecacl: A holistic framework for semi-supervised domain adaptation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. 8578– 8587, 2021

    2021

  22. [29]

    Semi- supervised domain adaptive structure learning,

    C. Qin, L. Wang, Q. Ma, Y . Yin, H. Wang, and Y . Fu, “Semi- supervised domain adaptive structure learning,” IEEE Transactions on Image Processing, vol. 31, pp. 7179–7190, 2022

    2022

  23. [30]

    Semi-supervised semantic seg- mentation with prototype-based consistency regularization,

    H.-M. Xu, L. Liu, Q. Bian, and Z. Yang, “Semi-supervised semantic seg- mentation with prototype-based consistency regularization,” Advances in neural information processing systems , 2022

    2022

  24. [31]

    Multi-level consistency learning for semi-supervised domain adaptation,

    Z. Yan, Y . Wu, G. Li, Y . Qin, X. Han, and S. Cui, “Multi-level consistency learning for semi-supervised domain adaptation,” arXiv preprint arXiv:2205.04066, 2022

    2022 arXiv

  25. [32]

    Semi-supervised domain adaptation with source label adaptation,

    Y .-C. Yu and H.-T. Lin, “Semi-supervised domain adaptation with source label adaptation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 24100–24109, 2023

    2023

  26. [33]

    Semi-supervised domain adap- tation via prototype-based multi-level learning,

    X. Huang, C. Zhu, and W. Chen, “Semi-supervised domain adap- tation via prototype-based multi-level learning,” arXiv preprint arXiv:2305.02693, 2023

    2023 arXiv

  27. [34]

    Adaptive betweenness clustering for semi- supervised domain adaptation,

    J. Li, G. Li, and Y . Yu, “Adaptive betweenness clustering for semi- supervised domain adaptation,” IEEE Transactions on Image Processing, 2023

    2023

  28. [35]

    Inter-domain mixup for semi-supervised domain adaptation,

    J. Li, G. Li, and Y . Yu, “Inter-domain mixup for semi-supervised domain adaptation,” Pattern Recognition, vol. 146, p. 110023, 2024

    2024

  29. [36]

    Semi-supervised domain adaptation for major depressive disorder detection,

    T. Chen, Y . Guo, S. Hao, and R. Hong, “Semi-supervised domain adaptation for major depressive disorder detection,” IEEE Transactions on Multimedia, 2023

    2023

  30. [37]

    Attract, perturb, and explore: Learning a feature alignment network for semi-supervised domain adaptation,

    T. Kim and C. Kim, “Attract, perturb, and explore: Learning a feature alignment network for semi-supervised domain adaptation,” in European conference on computer vision , pp. 591–607, Springer, 2020

    2020

  31. [38]

    Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation,

    J. Liang, D. Hu, and J. Feng, “Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation,” in International conference on machine learning , pp. 6028–6039, PMLR, 2020

    2020

  32. [39]

    Hard negative examples are hard, but useful,

    H. Xuan, A. Stylianou, X. Liu, and R. Pless, “Hard negative examples are hard, but useful,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16, pp. 126–142, Springer, 2020

    2020

  33. [40]

    Challenging tough samples in unsupervised domain adaptation,

    L. Zuo, M. Jing, J. Li, L. Zhu, K. Lu, and Y . Yang, “Challenging tough samples in unsupervised domain adaptation,” Pattern Recognition, vol. 110, p. 107540, 2021

    2021

  34. [41]

    Complementary attention-driven contrastive learning with hard-sample exploring for unsupervised do- main adaptive person re-id,

    Y . Liu, H. Ge, L. Sun, and Y . Hou, “Complementary attention-driven contrastive learning with hard-sample exploring for unsupervised do- main adaptive person re-id,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 1, pp. 326–341, 2022

    2022

  35. [42]

    Confidence- based visual dispersal for few-shot unsupervised domain adaptation,

    Y . Xiong, H. Chen, Z. Lin, S. Zhao, and G. Ding, “Confidence- based visual dispersal for few-shot unsupervised domain adaptation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 11621–11631, 2023

    2023

  36. [43]

    mixup: Beyond empirical risk minimization,

    H. Zhang, M. Cisse, Y . N. Dauphin, and D. Lopez-Paz, “mixup: Beyond empirical risk minimization,” arXiv preprint arXiv:1710.09412 , 2017

    2017 arXiv

  37. [44]

    Early- learning regularization prevents memorization of noisy labels,

    S. Liu, J. Niles-Weed, N. Razavian, and C. Fernandez-Granda, “Early- learning regularization prevents memorization of noisy labels,”Advances in neural information processing systems , vol. 33, pp. 20331–20342, 2020

    2020

  38. [45]

    When source-free domain adaptation meets learning with noisy labels,

    L. Yi, G. Xu, P. Xu, J. Li, R. Pu, C. Ling, A. I. McLeod, and B. Wang, “When source-free domain adaptation meets learning with noisy labels,” arXiv preprint arXiv:2301.13381 , 2023

    2023 arXiv

  39. [46]

    Multi-adversarial domain adaptation,

    Z. Pei, Z. Cao, M. Long, and J. Wang, “Multi-adversarial domain adaptation,” in Thirty-second AAAI conference on artificial intelligence , 2018

    2018

  40. [47]

    Deep hashing network for unsupervised domain adaptation,

    H. Venkateswara, J. Eusebio, S. Chakraborty, and S. Panchanathan, “Deep hashing network for unsupervised domain adaptation,” in Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5018–5027, 2017

    2017

  41. [48]

    Adapting visual cate- gory models to new domains,

    K. Saenko, B. Kulis, M. Fritz, and T. Darrell, “Adapting visual cate- gory models to new domains,” in Computer Vision–ECCV 2010: 11th European Conference on Computer Vision, Heraklion, Crete, Greece, September 5-11, 2010, Proceedings, Part IV 11 , pp. 213–226, Springer, 2010

    2010

  42. [49]

    A kernel two-sample test,

    A. Gretton, K. M. Borgwardt, M. J. Rasch, B. Sch ¨olkopf, and A. Smola, “A kernel two-sample test,” The Journal of Machine Learning Research, vol. 13, no. 1, pp. 723–773, 2012

    2012

  43. [50]

    Domain-adversarial training of neural networks,

    Y . Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Lavi- olette, M. Marchand, and V . Lempitsky, “Domain-adversarial training of neural networks,” The journal of machine learning research , vol. 17, no. 1, pp. 2096–2030, 2016

    2016

  44. [51]

    Deep transfer learning with joint adaptation networks,

    M. Long, H. Zhu, J. Wang, and M. I. Jordan, “Deep transfer learning with joint adaptation networks,” in International conference on machine learning, pp. 2208–2217, PMLR, 2017

    2017

  45. [52]

    Correlation alignment for unsupervised domain adaptation,

    B. Sun, J. Feng, and K. Saenko, “Correlation alignment for unsupervised domain adaptation,” Domain adaptation in computer vision applications, pp. 153–171, 2017

    2017

  46. [53]

    Deep unsupervised convolutional domain adaptation,

    J. Zhuo, S. Wang, W. Zhang, and Q. Huang, “Deep unsupervised convolutional domain adaptation,” in Proceedings of the 25th ACM international conference on Multimedia , pp. 261–269, 2017

    2017

  47. [54]

    Adversarial discrim- inative domain adaptation,

    E. Tzeng, J. Hoffman, K. Saenko, and T. Darrell, “Adversarial discrim- inative domain adaptation,” in Proceedings of the IEEE conference on computer vision and pattern recognition , pp. 7167–7176, 2017. 13

    2017

  48. [55]

    Joint distribution alignment via adversarial learning for domain adaptive object detection,

    B. Zhang, T. Chen, B. Wang, and R. Li, “Joint distribution alignment via adversarial learning for domain adaptive object detection,” IEEE Transactions on Multimedia , vol. 24, pp. 4102–4112, 2021

    2021

  49. [56]

    Learning semantic representa- tions for unsupervised domain adaptation,

    S. Xie, Z. Zheng, L. Chen, and C. Chen, “Learning semantic representa- tions for unsupervised domain adaptation,” in International conference on machine learning , pp. 5423–5432, PMLR, 2018

    2018

  50. [57]

    Wasserstein distance guided representation learning for domain adaptation,

    J. Shen, Y . Qu, W. Zhang, and Y . Yu, “Wasserstein distance guided representation learning for domain adaptation,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 32, 2018

    2018

  51. [58]

    Unsupervised domain adap- tation via deep conditional adaptation network,

    P. Ge, C.-X. Ren, X.-L. Xu, and H. Yan, “Unsupervised domain adap- tation via deep conditional adaptation network,” Pattern Recognition , vol. 134, p. 109088, 2023

    2023

  52. [59]

    Adversarial network with multiple classifiers for open set domain adaptation,

    T. Shermin, G. Lu, S. W. Teng, M. Murshed, and F. Sohel, “Adversarial network with multiple classifiers for open set domain adaptation,” IEEE Transactions on Multimedia , vol. 23, pp. 2732–2744, 2020

    2020

  53. [60]

    Progressive feature alignment for unsupervised domain adaptation,

    C. Chen, W. Xie, W. Huang, Y . Rong, X. Ding, Y . Huang, T. Xu, and J. Huang, “Progressive feature alignment for unsupervised domain adaptation,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pp. 627–636, 2019

    2019

  54. [61]

    Transferrable prototypical networks for unsupervised domain adaptation,

    Y . Pan, T. Yao, Y . Li, Y . Wang, C.-W. Ngo, and T. Mei, “Transferrable prototypical networks for unsupervised domain adaptation,” in Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 2239–2247, 2019

    2019

  55. [62]

    How does the combined risk affect the performance of unsupervised domain adaptation approaches?,

    L. Zhong, Z. Fang, F. Liu, J. Lu, B. Yuan, and G. Zhang, “How does the combined risk affect the performance of unsupervised domain adaptation approaches?,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, pp. 11079–11087, 2021

    2021

  56. [63]

    Improving semi-supervised domain adaptation using effective target selection and semantics,

    A. Singh, N. Doraiswamy, S. Takamuku, M. Bhalerao, T. Dutta, S. Biswas, A. Chepuri, B. Vengatesan, and N. Natori, “Improving semi-supervised domain adaptation using effective target selection and semantics,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Patt...

    2021

  57. [64]

    Deep co-training with task decomposition for semi-supervised domain adaptation,

    L. Yang, Y . Wang, M. Gao, A. Shrivastava, K. Q. Weinberger, W.-L. Chao, and S.-N. Lim, “Deep co-training with task decomposition for semi-supervised domain adaptation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. 8906–8916, 2021

    2021

  58. [65]

    Contradictory structure learning for semi-supervised domain adaptation,

    C. Qin, L. Wang, Q. Ma, Y . Yin, H. Wang, and Y . Fu, “Contradictory structure learning for semi-supervised domain adaptation,” in Proceed- ings of the 2021 SIAM International Conference on Data Mining (SDM), pp. 576–584, SIAM, 2021

    2021

  59. [66]

    Bidirectional adversarial training for semi-supervised domain adaptation.,

    P. Jiang, A. Wu, Y . Han, Y . Shao, M. Qi, and B. Li, “Bidirectional adversarial training for semi-supervised domain adaptation.,” in IJCAI, pp. 934–940, 2020

    2020

  60. [68]

    Representation learning with contrastive predictive coding,

    A. v. d. Oord, Y . Li, and O. Vinyals, “Representation learning with contrastive predictive coding,” arXiv preprint arXiv:1807.03748 , 2018

    2018 arXiv

  61. [69]

    Bootstrap your own latent-a new approach to self-supervised learning,

    J.-B. Grill, F. Strub, F. Altch ´e, C. Tallec, P. Richemond, E. Buchatskaya, C. Doersch, B. Avila Pires, Z. Guo, M. Gheshlaghi Azar, et al. , “Bootstrap your own latent-a new approach to self-supervised learning,” Advances in neural information processing systems , vol. 33, pp...

    2020

  62. [70]

    Supervised contrastive learn- ing,

    P. Khosla, P. Teterwak, C. Wang, A. Sarna, Y . Tian, P. Isola, A. Maschinot, C. Liu, and D. Krishnan, “Supervised contrastive learn- ing,” Advances in neural information processing systems , vol. 33, pp. 18661–18673, 2020

    2020

  63. [71]

    A simple framework for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in International conference on machine learning , pp. 1597–1607, PMLR, 2020

    2020

  64. [72]

    Probabilistic contrastive learning for domain adaptation,

    J. Li, Y . Zhang, Z. Wang, and K. Tu, “Probabilistic contrastive learning for domain adaptation,” arXiv preprint arXiv:2111.06021 , 2021

    2021 arXiv

  65. [73]

    Heterogeneous contrastive learning: Encoding spa- tial information for compact visual representations,

    X. Huo, L. Xie, L. Wei, X. Zhang, X. Chen, H. Li, Z. Yang, W. Zhou, H. Li, and Q. Tian, “Heterogeneous contrastive learning: Encoding spa- tial information for compact visual representations,” IEEE Transactions on Multimedia, vol. 24, pp. 4224–4235, 2021

    2021

  66. [74]

    Semi-supervised contrastive learning with similarity co-calibration,

    Y . Zhang, X. Zhang, J. Li, R. Qiu, H. Xu, and Q. Tian, “Semi-supervised contrastive learning with similarity co-calibration,” IEEE Transactions on Multimedia, 2022

    2022

  67. [75]

    Learning from different samples: A source-free framework for semi-supervised domain adapta- tion,

    X. Huang, C. Zhu, B. Zhang, and S. Zhang, “Learning from different samples: A source-free framework for semi-supervised domain adapta- tion,” 2024

    2024

  68. [76]

    Class-aware contrastive semi-supervised learn- ing,

    F. Yang, K. Wu, S. Zhang, G. Jiang, Y . Liu, F. Zheng, W. Zhang, C. Wang, and L. Zeng, “Class-aware contrastive semi-supervised learn- ing,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14421–14430, 2022

    2022

  69. [77]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pp. 9729–9738, 2020

    2020

  70. [78]

    Mixmatch: A holistic approach to semi-supervised learning,

    D. Berthelot, N. Carlini, I. Goodfellow, N. Papernot, A. Oliver, and C. A. Raffel, “Mixmatch: A holistic approach to semi-supervised learning,” Advances in neural information processing systems , vol. 32, 2019

    2019

  71. [79]

    How does mixup help with robustness and generalization?,

    L. Zhang, Z. Deng, K. Kawaguchi, A. Ghorbani, and J. Zou, “How does mixup help with robustness and generalization?,” arXiv preprint arXiv:2010.04819, 2020

    2010 arXiv

  72. [80]

    On mixup regu- larization,

    L. Carratino, M. Ciss ´e, R. Jenatton, and J.-P. Vert, “On mixup regu- larization,” The Journal of Machine Learning Research , vol. 23, no. 1, pp. 14632–14662, 2022

    2022

  73. [81]

    Prevalence of neural collapse during the terminal phase of deep learning training,

    V . Papyan, X. Han, and D. L. Donoho, “Prevalence of neural collapse during the terminal phase of deep learning training,” Proceedings of the National Academy of Sciences, vol. 117, no. 40, pp. 24652–24663, 2020

    2020

  74. [82]

    Proxymix: Proxy-based mixup training with label refinery for source-free domain adaptation,

    Y . Ding, L. Sheng, J. Liang, A. Zheng, and R. He, “Proxymix: Proxy-based mixup training with label refinery for source-free domain adaptation,” Neural Networks, vol. 167, pp. 92–103, 2023

    2023

  75. [83]

    Understanding and improving early stopping for learning with noisy labels,

    Y . Bai, E. Yang, B. Han, Y . Yang, J. Li, Y . Mao, G. Niu, and T. Liu, “Understanding and improving early stopping for learning with noisy labels,” Advances in Neural Information Processing Systems , vol. 34, pp. 24392–24403, 2021

    2021

  76. [84]

    Prestopping: How does early stopping help generalization against label noise?,

    H. Song, M. Kim, D. Park, and J.-G. Lee, “Prestopping: How does early stopping help generalization against label noise?,” 2019

    2019

  77. [85]

    Semi-supervised learning by entropy minimization,

    Y . Grandvalet and Y . Bengio, “Semi-supervised learning by entropy minimization,” Advances in neural information processing systems , vol. 17, 2004

    2004

  78. [86]

    Clda: Contrastive learning for semi-supervised domain adaptation,

    A. Singh, “Clda: Contrastive learning for semi-supervised domain adaptation,” in Advances in Neural Information Processing Systems (M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, eds.), vol. 34, pp. 5089–5101, Curran Associates, Inc., 2021

    2021

  79. [87]

    Moment matching for multi-source domain adaptation,

    X. Peng, Q. Bai, X. Xia, Z. Huang, K. Saenko, and B. Wang, “Moment matching for multi-source domain adaptation,” in Proceedings of the IEEE International Conference on Computer Vision , pp. 1406–1415, 2019

    2019

  80. [88]

    Very deep convolutional networks for large-scale image recognition,

    K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556 , 2014

    2014 arXiv

  81. [89]

    Randaugment: Practical automated data augmentation with a reduced search space,

    E. D. Cubuk, B. Zoph, J. Shlens, and Q. V . Le, “Randaugment: Practical automated data augmentation with a reduced search space,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops , pp. 702–703, 2020

    2020

  82. [90]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al. , “Pytorch: An imperative style, high-performance deep learning library,” Advances in neural information processing systems , vol. 32, 2019

    2019

  83. [91]

    Visualizing data using t-sne.,

    L. Van der Maaten and G. Hinton, “Visualizing data using t-sne.,” Journal of machine learning research , vol. 9, no. 11, 2008

    2008

  84. [92]

    Grad-cam: Visual explanations from deep networks via gradient-based localization,

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-cam: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the IEEE international conference on computer vision , pp. 618–626, 2017

    2017

  85. [93]

    Transferability vs. dis- criminability: Batch spectral penalization for adversarial domain adap- tation,

    X. Chen, S. Wang, M. Long, and J. Wang, “Transferability vs. dis- criminability: Batch spectral penalization for adversarial domain adap- tation,” in International conference on machine learning, pp. 1081–1090, PMLR, 2019

    2019

  86. [94]

    A collaborative alignment framework of transferable knowledge extraction for unsuper- vised domain adaptation,

    B. Xie, S. Li, F. Lv, C. H. Liu, G. Wang, and D. Wu, “A collaborative alignment framework of transferable knowledge extraction for unsuper- vised domain adaptation,” IEEE Transactions on Knowledge and Data Engineering, 2022

    2022

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.